Supercharging Digital Twins With Artificial Intelligence Tools

In parts one and two of this series, we’ve looked at how agentic artificial intelligence (AI) can streamline claims operations and the appeals and grievance process. Now we look at how AI, combined with digital twins, can enable health plans to fine tune how they can improve care coordination, population health management and cost containment.

Digital twins are virtual models of processes, systems, and/or physical objects—including people. A digital twin uses data gathered from sensors, databases, wearables, and applications to simulate how its real-world twin operates. Twins can be developed through process mining, which acts like an MRI for processes, using actual event data to reveal disconnects, manual activities, and handoffs that negatively impact outcomes and increase operating costs.

Using digital twins, health plans can model their members’ journeys, population health trends, care management workflows, and much more. Digital twins can also mirror potential futures. This means that health plans can use these models to simulate and assess the impact of new automations, innovative benefits, evidence-based care plans, and participation in pilot programs. By doing this, they can make informed decisions before dedicating time and resources to these initiatives.

AI capabilities make digital twins even more effective by amplifying their capabilities with the following attributes:

• Intelligence: Digital twins excel at capturing the essence of a process by gathering data from multiple sources. With the assistance of AI agents and other AI tools, we can explore what’s happening in real time as the process evolves. This rapid analysis allows health plans to quickly identify any process hiccups, upstream and downstream impacts, gaps in care, missing data essential for decision-making, or manual bottlenecks.

• Predictive capabilities: AI tools can extrapolate future outcomes based on current data. As a digital twin simulates a process, AI agents within it can forecast where gaps will occur, what trends will emerge in a specific population, which commercial plan members are most likely to transition to different plans, etc. Health plans will have more data and insights for anticipating and planning for market shifts.

• Continuous improvement: AI/Agentic AI can analyze the results of real-world actions taken based on a digital twin simulation. The AI can then refine the virtual twin so that future simulations are even more thorough.

Putting AI and Digital Twins to Work

Health plans have numerous ways to utilize AI-enhanced digital twins to enhance their processes, ultimately providing better member experiences through reduced inefficiencies. Below are some key use cases that leading health plans are exploring:

• Operational optimization: By simulating various scenarios with a twin, payers can identify bottlenecks in claims processing, call centers, and other key operations. AI agents can then take these insights and implement them across workflows to enhance efficiency. Operational changes made based on data are more effective and measurable.

• Enhanced risk stratification: Digital twins facilitate improved risk detection by combining various data sources, such as claims, clinical information, call center histories, social determinants of health, and data from member wearables or in-home devices. An AI model continuously observes the virtual twin, identifying patterns among members who face a heightened risk of worsening existing conditions or are nearing the development of a chronic condition. With the help of digital twins, health plans can simulate member progress across various care pathways, enabling them to establish more effective guidelines, implement more targeted interventions, and reduce preventable hospitalizations.

• Enhanced care coordination: Health plans can leverage digital twins to optimize and personalize care management strategies, ultimately enhancing patient outcomes. For example, by creating a digital twin using deidentified, aggregated data from a specific member cohort—such as Medicare Advantage patients with congestive heart failure—AI tools can analyze patterns in nonadherence, missed appointments, and emergency department visits. The AI can then predict which members are most likely to benefit from early intervention. By simulating various intervention strategies, the digital twin enables health plans to plan targeted, personalized care, ultimately improving member experiences and outcomes while reducing costs.

• Personalized member engagement: Generative AI can enable health plans to quickly generate member education and outreach programs. Digital twins can help assess these programs against virtual representations of member segments. AI agents can analyze results and help implement changes to make the engagement tactics as effective as possible.

• Enhanced population health management: AI tools can reveal patterns that indicate why and when members go to the emergency department, skip their medications, or neglect managing their diets. Digital twins can simulate which populations are at greatest risk for falling into these patterns, then simulate the impact of different interventions. These simulations further predict future health care utilization based on current trends and member characteristics. This allows payers to allocate resources and manage networks more efficiently and proactively.

• Enhanced cost modeling: Health plans can assess business and care initiatives, including bundled payment models, new business lines, nontraditional health care benefits, and more, using digital twins and AI. The virtual twin can simulate the impact of an initiative on a large scale before it is launched. AI agents and analytics can assist in performing cost-benefit analyses, giving health plans insights into which initiatives to pursue.

Implementing AI-Enhanced Digital Twins

Digital twins powered by AI/generative AI technology can model nearly any physical or virtual process, or a combination of both. That means the possibilities are vast. Health plans could potentially offer individual digital twins to specific members, helping them manage complex health conditions or achieve wellness goals.

However, before a health plan can develop new services based on AI-enhanced digital twins, it must first assess its data situation and consider these 3 critical prerequisites:

• Available data: Twins and AI need substantial structured and unstructured data to create accurate simulations and analyses. Multimodal generative AI can process a wide variety of structured and unstructured data. The performance of digital twins is optimized with real-time information.

• High-quality, comprehensive data: The accuracy of a digital twin simulation relies on the quality of the real-world data it uses. Plans should ensure that the data sources a twin will draw from are accurate and up to date. AI tools can assist by analyzing data stores and identifying gaps in critical data.;

• Secure data: Protecting members’ personal health information within the digital twin environment is essential not only for compliance but also to maintain member trust and encourage their willingness to adopt individualized health management offerings.

AI- and generative AI-powered digital twins are likely to support increasingly sophisticated applications and personalization. These could include the integration of genomic data for ultra-personalized risk prediction, real-time adjustments of benefit designs based on population health trends, and AI-driven scenario planning for rapid responses to market changes.

Exploring the current capabilities of AI-enhanced digital twins will provide health plans with a solid foundation for these and other future applications, while also gaining valuable abilities for fine-tuning care and population health management today.

About the Author

Deepan Vashi is the EVP & Head of Solutions for Health Plans and Healthcare Services at Firstsource with over 27 years of experience in health plan IT, business operations, and consulting. He is renowned for his expertise in developing member-centered digital solutions and building cross-functional teams to ensure successful implementation. In his role at Firstsource, he spearheads solutions and strategy for health plans, including Intelligent Back Office, Health Tech Services, and Platform-based Solutions (BPaaS). Deepan has extensive knowledge of innovative technologies such as Process Mining, Digital Twin, AI, and Blockchain.